AdministratorFile Attachment2000e526coverv05b.jpg

Practical Reasoning and Ethical Decision

What role does reason play in our actions? How do we know whether what we do is right? Can practical reasoning guide ethical judgment?

Practical Reasoning and Ethical Decision presents an account of practical reasoning as a process that can explain action, connect reasoning with intention, justify practical judgments, and provide a basis for ethical decisions.

The first part of the book is a detailed critical overview of the influential theories of practical reasoning found in Aristotle, Hume, and Kant. The second part examines practical reasoning in the light of important topics in moral psychologyweakness of will, self-deception, rationalization, and others. The third part describes the role of moral principles in practical reasoning and clarifies the way practical reasoning underlies ethical decisions. Audi formulates a comprehensive set of concrete ethical principles, explains how they apply to reasoning about what to do, and shows how practical reasoning guides moral conduct.

Practical Reasoning and Ethical Decision provides the most comprehensive account of the topic in the current literature and is essential reading for anyone interested in the role of reason in ethics or the nature of human action.

Robert Audi is the David E.Gallo Chair in Ethics and Professor of Philosophy at the University of Notre Dame. He is the author of Moral Knowledge and Ethical Character (1997), The Architecture of Reason (2001), Epistemology (Routledge, 2003), The Good in the Right (2004), and many papers in ethics, moral psychology, and the philosophy of action.

Robert Audi

Practical Reasoning and Ethical Decision

LONDON AND NEW YORK

First published 2006 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX 14 4RN

Simultaneously published in the USA and Canada by Routledge 270 Madison Avenue, New York, NY 10016, USA

Routledge is an imprint of the Taylor & Francis Group This edition published in the Taylor & Francis e-Library, 2006.

To purchase your own copy of this or any of Taylor & Francis or Routledges collection of thousands of eBooks please go to http://www.ebookstore.tandf.co.uk/.

2006 Robert Audi

All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including

photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers.

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

Library of Congress Cataloging in Publication Data Audi, Robert, 1941 Practical reasoning and ethical decision/Robert Audi.1 st ed. p. cm. Includes bibliographical references and index.

ISBN 0-415-36462-0 (hardcover: alk paper)ISBN 0-415-36463-9 (pbk: alk paper) I. Practical reason. 2. Ethics. 3. Decision making. 4. Aristotle. 5. Hume, David, 17111776. 6. Kant,

Immanuel, 17241804. 1.Title. BC177.A84 2005 I70.42dc22 2005018550

ISBN 0-203-01568-1 Master e-book ISBN

ISBN 10: 0-415-36462-0 (Print Edition) ISBN 13: 978-0-415-36462-1 (hbk) ISBN 10: 0-415-36463-9 (Print Edition) ISBN 13: 978-0-415-36463-8 (pbk)

Taylor & Francis Group is the Academic Division of T&F Informa p/c.

Contents

Preface viii

Acknowledgments x

Introduction 1

PART I Historical and conceptual background: practical reasoning in Aristotle, Hume, and Kant

8

1 Aristotle on practical reasoning and the structure of action 10 2 Hume and the instrumentalist conception of practical reasoning 29 3 Kant and the autonomy of practical reason 45

Part II Practical reasoning, practical arguments, and intentional action 62

4 The varieties and basic elements of practical reasoning 64 5 Practical reasoning and intentional action 82 6 Practical reasoning in the dynamics of action 95

Part III Practical reasoning, ethical decision, and rational action 107

7 The assessment of practical reasoning 109 8 General principles of practical appraisal 124 9 Practical reasoning and moral judgment 135

10 Practical reasoning in ethical decisions 147 11 The rationality of action and the plurality of value 157

Conclusion 170

Notes 175

Index 194

Preface

This book is intended as a wide-ranging contribution to moral psychologyconceived as including the theory of actionand normative ethics. It is a sequel to Practical Reasoning, published by Routledge in 1989. It contains three new chapters that connect my theory of practical reasoning with ethics, and the content of the earlier volume has been expanded in many places and revisedoften in major waysin part with the idea of placing my account of practical reasoning in the context of an ethical theory that I have developed in The Good in the Right (Princeton, 2004) and elsewhere. Most of the content of the earlier, much shorter book is preserved, but some points have been eliminated. Many of the revisions reflect my thinking and writings in ethics and the theory of practical reason since the earlier book, and Chapter 8 draws heavily on my Reasons, Practical Reason, and Practical Reasoning (Ratio 17, 2, 2004).

In many places the revised chapters take into account various developments in ethics and the philosophy of action during the past fifteen years. This book is not, however, about major figures in the literature, apart from Aristotle, Hume, and Kant. It is intended as an independent contribution to moral psychology and normative ethics. Even in exploring their positions, then, I have concentrated on what seems philosophically central rather than on historical or textual questions.

Readers in either ethics or the philosophy of action will find many discussions in the book that bear on central problems in those fields. Among these are the moral standards that govern practical reasoning in ethical matters; the status of moral reasons in comparison with other practical reasons; the dimensions of reasoned ethical decisions; the structure and explanation of action, and the conditions for its rationality; the nature and dynamics of weakness of will; and the connection between practical reasoning and such psychological notions as inference, self-deception, rationalization, and unconscious motivation.

Readers interested in the history of ethics will find interpretations of major elements in the three immensely influential positions explored in Part I: Aristotles, which remains central in virtue ethics; Humes, which is central for understanding instrumentalismand, in my view, naturalismin the theory of practical reason; and Kants, which is currently the leading deontological view in modern ethics. These three positions, both as they bear on practical reasoning and in their implications for ethics, are kept in view throughout the book.

The writing style and narrative elements of the book are meant to make it accessible to serious students in any of the areas just mentioned. The aim has been to combine adequate clarity and concreteness to help readers coming to the topic for the first time with sufficient rigor and originality to reward professional readers.

For those with background in the subject, any of the three parts may be read separately, but this applies to Part I for virtually any reader. Most of the references to Part I that come later in the book are quite intelligible in their context or easy to pursue using

the index. Readers particularly interested in normative questions and ethical decision should find Part III largely self-contained. Indeed, Chapters 8 through 11 (or even 8 or 9 through 10) are largely understandable on their own. The Introduction provides a sense of the scope of all three parts; each chapter has a summary of its major points; and the Conclusion draws together main ideas developed throughout the book.

Acknowledgments

Given how much of its predecessor is preserved in this book, it is appropriate to note my gratitude to those who helped me in that earlier project: Richard Lee, Raimo Tuomela, Michael Zimmerman, and, especially, Hugh McCann and Alfred Mele. Each commented on the draft that preceded the 1989 publication. For comments during the same period I am grateful to Nelson Potter for comments on the first three chapters, Ralf Meerbote for helpful remarks on Chapter 3, and the late Robert Hurlbutt for detailed discussion of Chapter 2. I benefited much from conversations with Karl Ameriks on Chapter 3, David OConnor on Chapter 1, and W.David Solomon on Chapter 2.

I particularly want to acknowledge again the value of conversations I have had, over a number of years, on practical reasoning and related topics, with William Alston, Michael Bratman, Carl Ginet, Hugh McCann, Alfred Mele, and Raimo Tuomela. I regret that space constraints made it impossible to take adequate account of their work. On one or another topic in the book, I have profited from discussions or correspondence with Frederick Adams, Bruce Aune, Lewis White Beck, Myles Brand, Richard Brandt, Paul Churchland, Norman Dahl, Stephen Darwall, Richard Foley, William Frankena, Brian McLaughlin, Alvin Plantinga, Amlie Oksenberg Rorty, John Searle, James Sterba, Judith Thomson, and Peter Vallentyne. Numerous discussions with my students have also contributed much to my thinking on topics addressed in the book.

Since 1989 I have also benefited from conversations on practical reason or related ethical topics with many philosophers, including some in audiences for papers I have given. I cannot list them all but would particularly like to mention Julia Annas, John Broome, Roger Crisp, Garrett Cullity, Jonathan Dancy, James Dreier, John Fischer, Richard Fumerton, Berys Gaut, Bernard Gert, Allan Gibbard, John Greco, Gilbert Harman, Brad Hooker, Philip Kain, Jaegwon Kim, Christine Korsgaard, Christopher Kulp, Scott LaBarge, Noah Lemos, Alasdair MacIntyre, Joseph Mendola, Michael Meyer, Christian Miller, Elijah Millgram, Thomas Nagel, Mark Nelson, Martha Nussbaum, Onora ONeill, Derek Parfit, John Perry, William Prior, Elizabeth Radcliffe, Bruce Russell, Geoffrey Sayre-McCord, Tim Scanlon, Russ Shafer-Landau, Walter Sinnott-Armstrong, Michael Smith, Robert Solomon, Ernest Sosa, Philip Stratton-Lake, Eleonore Stump, Mark Timmons, William Tolhurst, Mark van Roojen, R.Jay Wallace, Ralph Wedgwood, and Michael Zimmerman. Roger Crisp, Derek Parfit and Sean Patrick Walsh generously provided helpful comments on a draft of the entire book.

For advice and support at many points I am grateful to Tony Bruce, Lucie Ewin, Sonja van Leeuwen, and, for copyediting, Tony Nixon.

Introduction

Human life is pervaded by problems. Reasoning is a common response to problems we care about. We do it more than we notice; it is natural, often automatic, and need not call attention to itself. Reasoning is often associated with intellectual problems, such as what constitutes a just state, or whether a system of axioms is free of contradiction. But we also reason in response to everyday problems about how to get on with ordinary business. Confronted with a request for help at a time when we have planned something else, we must decide whether to decline or to adjust our schedule; given a roadblock, we must choose among detours.

Reasoning appropriate to problems of the first kind has often been called theoretical; reasoning appropriate to problems of the second kind has often been called practical. These terms do not imply that every theoretical problem is connected with an actual theory, such as a philosophical or scientific one, or that every practical problem is unrelated to theories. Indeed, a theoretical problem may be as routine as determining the cause of a fire, and a practical problem may be as scientific as constructing a safe tunnel. Historically, the main point of the terminology is to suggest that practical problems are addressed to us as agents and concern what we are to do, whereas theoretical problems are addressed to us as knowers, or potential knowers, and concern questions of what is (or is not) true. The former problems are solved by practice, say by taking the right detour; the latter are solved by theory, typically by our forming, or bringing to bear, the right belief, say by our constructing a sound proof that an axiom system is consistent.

Another way to draw the contrast is through distinguishing between practical and theoretical reasons. Practical reasons might be said to be reasons for acting; theoretical reasons might be described as reasons for believing. If, for instance, I have promised to attend a certain conference in London, then I have, and my want to attend it expresses, a practical reason for going to London. If, accordingly, I go to London in order to attend it, my going will be done at least partly for that reason: to attend. On the other hand, suppose I believe that the topic is Humes ethics. I have, and my belief expresses, a prima facie theoretical reason for believing that Humes idea of reason as the slave of the passions will be discussed. If I believe the latter proposition on the basis of the former, I believe it at least partly for the reason in question: namely, that the topic is Humes ethics.

Perhaps we might say that whereas theoretical reasons in some sense point toward truth, practical reasons in some sense point toward action. The idea, in part, is that a good theoretical reason in some way supports a proposition for which it is a good reason and also supports, though it need not entail, the rationality of believing any such proposition; a good practical reason supports an action for which it is a good reason, and also supports, though it need not entail, the rationality of performing the action for that reason.

It is clear that desires can expressin the sense that their content can constitutepractical reasons. Avoiding a dispute with a friend may be what I want to achieve in a delicate situation and may be my reason for not raising a problem that divides us. But beliefswhich, unlike desires, are true or false and might be thought to be theoretical attitudescan also express practical reasons. Consider the belief that I would avoid a dispute with a friend if I did not mention the impending divorce of a colleague. This proposition is a clear case of one I might cite if asked my reason for avoiding that topic for an entire evening among mutual friends concerned about the likely break-up. Taking beliefs to express practical reasons leaves open whether they can do this only by indicating something other than the proposition believed, such as a goal of the action in question. Whether this is so or not, beliefs as well as desires figure in expressions of the reasons crucial for practical reasoning and ethical decision, which is the central topic of this book, and both beliefs and desires will be examined in some detail.

A related problem on which the book bears (though it too will be discussed only in relation to practical reasoning) is whether practical and theoretical reasons may be objective, in the sense that they may be reasons for anyone to act or to believe accordingly, or just subjective, in the sense that they are reasons only for a particular person. That a conference on Humes ethics will be in London may be a reason for me to go there; it is not a reason for just anyone to do so. But perhaps an actions being cruel is a reason for anyone to abstain from it. Moreover, that a conference is on Humes ethics perhaps is a reason for anyone, or at least anyone generally informed about Hume, to believe that reason and passion will be discussed, though of course one could have such an objective reason without realizing one does.

Practical and theoretical reasoning have both been much discussed by philosophers, but the latter has had a larger share of their attention. It is often treated, if only implicitly, in the course of teaching logic; it is stressed in many epistemological works; and it figures centrally in numerous discussions in the philosophy of science, metaphysics, and other areas of philosophic inquiry. This book mainly concerns practical reasoning, especially in contexts of moral decision, but practical reasoning can be adequately understood only in relation to theoretical reasoning. Indeed, a satisfactory account of practical reasoning must enable us to see both why it is appropriately called reasoning and how it is related to theoretical reasoning. At several points, then, I will discuss similarities and differences between these two kinds of reasoning.

A special problem we face in discussing practical reasoning is that no everyday expression can guide us in clarifying it, in the way terms like explanation, knowledge, and justice can guide us in understanding the concepts they express. As philosophers use practical reasoning, it islike practical reason and theoretical reasona term of art. It has little life in ordinary parlance and a multiple personality in philosophical literature. Fortunately, there are terms crucial for understanding practical reasoning, such as reasoning, inferring, and concluding, which do have sufficiently settled uses to help us focus certain of our ideas. Furthermore, there are several problems that, in discussing the nature and role of practical reasoning, philosophers have wanted to solve. We can also use these problems in guiding our inquiry, and I will discuss them in detail.

To get a better sense of the problems to be addressed by an account of practical reasoning, consider an example. Suppose I enter my living room and find two guests, a husband and wife, quarreling. I am disturbed. I like them both, and I have a problem: how

Practical reasoning and ethical decision 2

to reconcile them, or at least temporarily make peace between them. I listen to see if there is simply a misunderstanding. I find the matter more complex: she, Janet, has criticized him, William, for not making time to do something with their children on the weekend, and they are beginning to argue about who is responsible for various aspects of the childrens activities on the coming weekend. It seems that they each made weekend plans on the assumption that the other would take over the children, and both are trying to make minimal adjustments. I begin wondering how I can help.

It now occurs to me that I could do something with their children and mine, thereby helping with the problem. I think about this prospect a moment and become convinced that I can manage all the children. I come to believe that offering to do something with all the children this weekend would help end the quarrel, and I conclude that I should offer this. Though slightly restrained by my awareness of all the extra effort, I quickly decide to propose my plan. I wait a bit for a good time to interrupt them, and I then suggest that I might help this weekend by getting all the children together at my home. Making this suggestion is my (attempted) solution: I am trying to resolve the reconciliation problem by removing an obstacle to their harmony.

The example shows a number of important elements in what we might broadly call practical thinking, using this term to encompass not only practical reasoning, but also less structured reflection on the same problem to whose solution the practical reasoning is directed. First, I am confronted by a problem of what to do. It is a (felt) problem for me because of what I care about: very roughly, there is something I want, here harmony between friends, and I take it that I cannot have it unless I do something appropriate, or at least I suppose that I may be able to get it by doing the right thing. Second, I deliberate about how to achieve what I want, a process which includes such things as reflecting on apparent options, recalling facts, and reasoning theoretically. For instance from the supposition that I take their children over on the weekend, I may conclude that telling William and Janet that I will do so would likely remove an obstacle to their harmony. Thirdand perhaps simultaneouslyI form the belief that offering to take the children over would help end the quarrel. Fourth, I judge, partly on the basis of this belief, that I should offer. And finally, after waiting for a good opportunity, I act on this judgment: I suggest to Janet and William that I get all our children together.

If, in a stretch of such practical thinking, we try to isolate something that it is natural to call practical reasoning, there need be no one piece of reasoning which clearly fills this role. Suppose, however, we are guided by the thought that we should identify a piece of reasoning that is a response to a practical problem and concludes with an answer to the problemin the sense of an answer to the associated question, which is here roughly this: How can I reconcile Janet and William? The best candidate is probably my reasoning from the premise that I want to reconcile them, together with the related premise (itself arrived at by theoretical reasoning) that suggesting I take over all our children this weekend will reconcile them, to the conclusion that I should suggest it. I have reasoned to an answer to my problem; the answer is practical in the sense that it indicates an action of mine which it represents as a means of solving my practical problem; and my acting in accordance with that answer is my attempted solution.

In different terminology, my reasoning both arises from a motivational state expressing a practical reasonfrom my wanting to reconcile my friendsand generates another practical reason, by concluding with the judgment that I should suggest taking

Introduction 3

over the children. Finally, my acting on that reasonby making this suggestionis what I take to be my solution to my problem. In speaking of generating a practical reason, we need not deny that there may have already been an objective reason to do what I judged I should do; but until I in some way become aware of this reason, it will not be, for me, a practical reason that motivates my solving my problem.

There are, however, many other strategies for identifying practical reasoning, and in examining some of what philosophers have said about such reasoning I want to compare other strategies with the one just illustrated. If practical reasoning had a sufficiently settled use, then one good way of choosing among the various accounts of such reasoning might be to appeal to intuitions about the application of the term. In the absence of such a use, we need a different way to ascertain what constitutes practical reasoning.

As it happens, there are certain major problems that give the topic of practical reasoning its special interest. These are philosophically important in their own right, often in relation to making decisions in ethical matters. But one or another of them has also been a main concern of philosophers writing on practical reasoning, and all seem discernible in Aristotle, who is, historically, the most important writer on the topic. This book aims above all at developing at least partial solutions to these problems. If this end is achieved, then even if the specific account of practical reasoning that is offered does not capture all the plausible conceptions of it found in the philosophical literature, we shall at least have a framework for understanding those conceptions.

The problems that seem central to the topic of practical reasoning can all be illustrated with respect to our reconciliation example. Let me briefly sketch each one.

The first problem is to connect practical reasoning with the kinds of questions that provide the occasion for it. Above all, how is a practical question rationally answered? The problem is to give at least a partial account of how a rational personsay the kind Aristotle called a person of practical wisdomanswers a practical question, such as What am I to do to reconcile Janet and William? Clearly, practical reasoning, as described above, constitutes one way that a rational person produces an answer. I take it that such a question can arise for an agent without being asked, whether by that agent or anyone else. Simply through confronting a practical problem, one can feel the need to answer a practical question. Thus, the occasions that call for practical reasoning go far beyond cases in which a practical question is explicitly asked. The context may be social, as in our example, or personal, as where one is planning a quiet afternoon. The question may be instrumental, say about how best to make up for time lost in a detour; or it may concern intrinsic ends, for instance when, with only enjoyment in view, one considers which of several books to read on a weekend. And ones options may, like alternative vacation trips, call for complex planning, or they may arise spontaneously, as where we are offered a choice of desserts, or come upon two equally appealing paths on a woodland walk.

The second problem centers on the question, What is it for an agent to act for a reason? Since reasons on which one acts are (in at least a minimal sense) practicaland any practical reason can in some way be acted onthis is roughly the question of how a practical reason grounds an action based upon it. Clearly, I act for a reason when, in order to help reconcile Janet and William, I suggest my taking over all the children on the weekend. Acting on the basis of practical reasoning is a paradigm of acting for a reason. It may indeed turn out that either actions so based are the only ones performed for a

Practical reasoning and ethical decision 4

reason, or, at least, in giving an account of action based on practical reasoning, we can exhibit the crucial elements in acting simply for a reason. The central question here is whether, to some extent at least, action for a reason can be explicated as action based on practical reasoning. More broadly still, to what degree do reasons for which we act operate through reasoning?

The third problem is structural. Is there a pattern of reasoning by which all intentional actions can be explained, including actions exhibiting weakness of will? (I take actions exhibiting weakness of will to beroughlyuncompelled, and normally intentional, actions against ones better judgment.) There are two main difficulties here. One is to ascertain how reasoning processes might figure in producing our actions. Call this the explanatory problem. The other is to determine whether all intentional action is intelligible in terms of a pattern that applies to the agents psychological state at the crucial time, whether or not this pattern is expressed in a reasoning process. Call this the intelligibility problem. Suppose that my desire to avoid extra work had prevailed over my judgment that I should suggest taking over all of the children on the weekend. By keeping silent and thereby failing to act on that judgment, I might have exhibited weakness of will. We need an account of how the motivational pattern causing me to keep silent can interact with the pattern of motivation and judgment corresponding to my practical reasoning. How, in a generally rational person, can weakness of will prevail over a conflicting judgment regarding what one should do?

Suppose, on the other hand, that I had not reasoned my way to the judgment that I should take over all the children, yet simply made that judgment as an immediate intuitive response to my problem. Both the judgment and my action based on it might still be intelligible in relation to the motivational and cognitive elements (roughly, belief elements) that are represented in practical reasoning. The same kind of intelligibility apparently characterizes what underlies the imagined weak-willed failure to act on my judgment: the relevant omission is intentional and seems attributable to a similar (prevailing) pattern. As these cases suggest, our structural problem is both to ascertain the explanatory role of practical reasoning and to determine how the patterns it exhibits may render intentional actions intelligible.

By contrast with the third problem, the fourth concerns the dynamics of action: the events or processes or whatever that produce an intentional action, given the agents reasons for it. Specifically, what is it that causally mediates between practical reasons and actions? The question here is what it is that causally links ones motives, intentions, beliefs, and other psychological elements that express ones reasons to act, and, on the other hand, the actions one performs for those reasons. Consider my (rationally) wanting to reconcile Janet and William, and thereby having a reason to act toward that end. This motivational state does not automatically yield such action; I may, for instance, have no thought which points toward an available means to accomplish my end. Practical reasoning that concludes in favor of a specific action seems to fill the bill perfectly: it is a process with both the right sorts of constituents to motivate and direct the action, and the right kind of content and conscious manifestations to trigger the action. For instance, in some way the reasoning embodies adequate motivation, and it concludes with a judgment that I should suggest taking over all the children.

The fifth problem concerns the rationality of action. Can we so specify the structure of acting for a reason that actions for a reason can be seen, in the light of that reason, as not

Introduction 5

only intelligible but also prima facie rational: roughly, rational apart from certain defeating conditions, such as may occur in certain cases of weakness of will? If actions for a reason are always based on practical reasoning, or are even somehow undergirded by a motivational and cognitive pattern of the kind that yields practical reasoning, then it seems plain why they should be (at least weakly) prima facie rational: they are grounded in reasoning and to that extent guided by reason. This is especially likely to seem a plausible explanation of their prima facie rationality if we presuppose a notion of practical reasoning which, as in our example, makes it minimally acceptable by appropriate standards for evaluating the cogency of reasoning. Suppose my reasoning proceeds from (1) my positing an end not overridden by any competing end of mine, and (2) my reasonable belief that doing a certain thing will easily accomplish that end, to (3) the conclusion that I should do it. If, for instance, I presently want to reconcile Janet and William more than to do anything that might compete with this aim, and I believe (for adequate reason) that offering to take over the children will accomplish my end, then judging that I should offer is a prima facie quite reasonable inference to draw, making the offer is a prima facie quite reasonable thing to do, and actually offering to do this is a prima facie rational concrete action.

The last problem I want to stress concerns the degree of unity of the notion of reasoning in general. Specifically, how can we account for both the similarities and the differences between practical and theoretical reasoning? In virtue of what, for instance, are they both reasoning? Do they differ in relation to the standards appropriate to their evaluation, or even in respect to what sorts of items constitute their premises and conclusions? Or do they differ mainly in the kinds of reasons they give us, or perhaps simply in the sorts of problems that motivate our engaging in them? Here, too, there is a considerable diversity of views in the philosophical literature.

If I were to add another problem giving special interest to the topic of practical reasoning, it might be to understand Aristotles views on practical thinking and the nature of human action. Aristotle appears again and again in discussions of practical reasoning, especially in relation to ethical questions, and neither the history of the topic nor the current literature can be understood wholly in abstraction from his work. It is appropriate, then, to begin a study of practical reasoning with him. It is also appropriate to explore other important historical figures, though this book is not historical and is concerned chiefly with the six general questions just raised and their significance in ethics. In the literature that addresses or bears on practical reasoning, Hume and Kant are, after Aristotle, perhaps the most important. They are certainly major figures in this literature; each contrasts sharply with the other; and each has significant similarities to Aristotle, though they both depart from him in major ways. There are, of course, other important historical figures I simply cannot take up. My hope is that if they are as important for the topic as Hume or Kant, their positions can either be seen as, in good part, combinations of elements of those of Aristotle, Hume, and Kant, or at least as understandable largely on the basis of the treatment of practical reasoning and ethical decision which, beginning with those three philosophers, this book will give.

Practical reasoning and ethical decision 6

Part I Historical and conceptual

background Practical reasoning in Aristotle, Hume, and

Kant

One Aristotle on practical reasoning and the

structure of action

Aristotles principal writings on practical thinking are in the Nicomachean Ethics (NE), and this will be my main Aristotelian text, though some references will be made to others of his works. The Ethics is a densely packed, rich text, and I cannot hope to formulate the Aristotelian position on practical reasoning, if there is just one such position. Nor can I even begin to do justice to the large body of valuable literature on Aristotles account of practical reasoning. My aim is simply to formulate and interpret one plausible Aristotelian conception (or range of conceptions) of practical reasoning, particularly as it appears in the Ethics, and to identify, through exploring Aristotle, some major concepts and problems crucial for understanding practical reasoning in general.

Aristotle did not use any term that can be literally translated as practical reasoning,1 and what he called practical syllogisms may represent a narrower category than what I have so far called practical reasoning. In his treatment of practical thinking in general, however, he said a great deal about deliberation, which some commentators take him to equate with practical reasoning.2 Moreover, deliberation is plainly a kind of practical thinking; and, as illustrated by the reconciliation example in the Introduction, deliberation may embody what I have called practical reasoning. It may be wise, then, to begin with Aristotles views about deliberation and work from what we learn there toward an understanding of his views of practical reasoning.

1 DELIBERATION

Regarding the objects of deliberation, Aristotle makes both positive and negative points:

[W]e deliberate about what results through our own agency, but in different ways on different occasions, e.g. about questions of medicine and money-making. We deliberate not about ends, but about what promotes ends; a doctor, e.g., does not deliberate about whether he will cure, or an orator about whether he will persuade, or a politician about whether he will produce good order, or any other [expert] about the end [that his science aims at]. Rather, we first lay down the end, and then examine the ways and means to achieve it. If it appears that any of several [possible] means will reach it, we consider which of them will reach it most easily and most finely; and if only one [possible] means reaches it,

we consider how that means will reach it, and how the means itself is reached, until we come to the first cause, the last thing to be discovered.3

To illustrate with one of Aristotles examples, if I am a physician treating a patient, my governing end as physician (leaving aside the issue of euthanasia) is to cure, and I do not deliberate about whether I will (or should) cure the patient. I do, however, deliberate about means, say about whether I should give medicine or simply recommend rest.

At least two conceptions of deliberation are consistent with this passage. On one, the deliberative chain contains a series of decisions leading to the final decision which is, or is at least closely tied to, the first thing in the order of causation. In this first instance, the deliberative chain is decisional: if I decide to prescribe medicine, doing so becomes a subsidiary end, and I may then deliberate about what medicine I should give. If I decide on penicillin as a means of cure, I have another subsidiary end and may deliberate about how I should carry that out, say by tablet or injection. If I now decide on tablets, I may realize that they are in the cabinet to my left. Suppose I decide to give some of those very tablets; then, aware that I need only reach for them, I do it. On the second conception consistent with the passage, although I make the same final decision, the deliberative chain is cognitive: instrumental beliefs (or other cognitive elements, such as judgments) express the subsidiary ends; for instance, I do not decide to prescribe medicine, but do judge prescribing it to be best and thereby proceed to identify the best medicine, and then the best vehicle for giving it. I finally decide to do the thing that is warranted by the entire sequence: reaching for the tablets.

Schematically, the difference between the two kinds of chain is the kind that exists between (1) deciding to A, which one believes one can well achieve by B-ing, deciding to B, which one believes one can well achieve by C-ing, and so on until one decides on something here and now, such as reaching for the tablets, and (2) forming the beliefs that A can be well achieved by B-ing, that B-ing can be well achieved by C-ing, and so on until one reaches something one can do here and now, which one decides to do. In both kinds of chain there will be appropriate instrumental beliefs. They are in fact required to explain the subsidiary decisions. But in one kind of chain there are subsidiary decisions; in the other, not.

The decisional interpretation of the chain may be more often what Aristotle had in mind in speaking of such chains. Moreover, it is quite consistent with his overall views to allow cases in which the decisions are conditional. There, when one reaches the end, action will follow only on a further condition; for instance, if one decides to give tablets provided there are enough, then one would check before giving them, and give them only if one finds enough. I prefer the decisional interpretation for most of the relevant passages; but the more economical, cognitive reading may better fit others. In any case, no major point below turns on which interpretation is taken.

The descriptions just given seem to encompass the completion of the process of deliberation, but they do not indicate what, exactly, is the first link in the chain of causation. If that is the last step in the order of discovery, one would think it is my final means to realizing my end, say reaching for the tablets. For this is the final means I take to be necessary, and it seems to originate the causal chain leading (if I succeed) to a cure. But the text is not without vagueness on this matter:

Aristotle on practical reasoning and the structure of action 11

What we deliberate about is the same as what we decide to do, except that by the time we decide to do it, it is definite; for what we decide to do is what we have judged [to be right] as a result of deliberation. For each of us stops inquiring how to act as soon as he traces the origin to himself, and within himself to the dominant part; for this is the part that decides. We have found, then, that what we decide to do is whatever action among those up to us we deliberate about and desire to do. Hence also decision will be deliberative desire to do an action that is up to us; for when we have judged [that it is right] as a result of deliberation, our desire to do it expresses our wish.

(NE: 1113a212)

If decision is deliberative desire, how can it be identical with something that is, in an active way, made, as decisions are? A decision is, if not an action, at least an event, whereas desires are not events, in the usual sense in which the occurrence of an event entails that of change (the existence of a desire does not entail that of change). One possible answer may be that deliberative desire should be taken to be (in English) a technical term, and the weight should be put on decision, which does appear to designate action or at least behavior (the Greek term in question, prohairesis, is also commonly translated choice, which confirms this actional reading).4 Decision, moreover, entails desire. The decision would be deliberative by virtue of its rootedness in a deliberative process; it would be conative by virtue of expressing our wish. If we add that it is an active expression of that wish, we may think of it as action-like.

Such an interpretation may be developed in at least two different directions. First, in cases like this, in which we deliberatively reach, and immediately perform, a bodily action in our power as the final means to our overall end, Aristotle may have thought of the decision to do the thing in question and the doing of it as the same action under two different descriptions. One need not decide to reach for the tablets and then do so; ones deciding to give them to the patient coincides with reaching for them; it occurs straightaway upon ones realizing that the tablets are in the cabinet on ones left. If this is correct, then a patient who was aware of the chain of deliberation and thus said, observing ones taking them from the cabinet, I see you decided to give me the tablets, would be saying nothing beyond what the passage licenses, and would preserve the vagueness of the reference of decide. On the other hand, we might instead suppose that decision is, or is a precursor of, volition, understood as an act of will.5

To be sure, we might treat volition as ordinary action under a special kind of description, and in that case this interpretation would be quite similar to the first. But volition is more commonly taken to be a kind of doing that is not action, or as a sort of active intending to do something here and now.6 In what follows, we need not choose either interpretation; indeed my main points about Aristotle concern what happens prior to decision and will in any event be consistent with both interpretations, and indeed with taking decision to be a volitional state as opposed to an event of any kind.

Practical reasoning and ethical decision 12

2 THE PRACTICAL SYLLOGISM

If we can locate what Aristotle called the practical syllogism in relation to his deliberative chains, this may give us a better idea of the sort of reasoning he conceived as practical in the sense sketched in the Introduction. In discussing weakness of will (which will be considered shortly) Aristotle said much about the practical syllogism. Here is an important passage about such syllogisms in general:

One belief (a) is universal; the other (b) is about particulars, and because they are particulars perception controls them. And in the cases where these two beliefs result in (c) one belief, it is necessary in purely theoretical beliefs for the soul to affirm what has been concluded, and in beliefs about production (d) to act at once on what has been concluded.

If, e.g., (a) everything sweet must be tasted, and (b) this, some one particular thing, is sweet, it is necessary (d) for someone who is able and unhindered also to act on this at the same time,

(NE: 1147a2531)

where (c), the one belief that is a result of (a) and (b), is presumably the belief that this must be tasted. If we now recall the deliberative chain leading to reaching for tablets to cure the patient, we might take, as a clue in locating a practical syllogism like this one in the chain, the closeness of the syllogism to actiona feature of such syllogisms which Aristotle emphasizes elsewhere too.7 Perhaps such a practical syllogism begins when I reach the conclusion (from previous reasoning) that diseases of the sort this patient has are to be treated with penicillin tablets. With this goal in view, I realize, perceptually, that they are in the cabinet to my left and judge that I must give them. I am bound to act on this at the same time.8

It is noteworthy that Aristotle speaks in this passage both of the souls being bound to affirm the conclusionwhich in the example we are considering would be that this must be tasted [by me]and of the agent s being bound to perform this act at once. If the act is tasting, we need a behavioral referent to make sense of what Aristotle is saying; if it is the souls affirming the conclusion, we need a mental referent. The view that the crucial action is a decision (or a choice) conceived as also capable of bearing a physical behavioral description gives us precisely what we need. But Aristotle rightly refers to each category in distinct terms, since the crucial decision need not bear a physical act description, say whereas Aristotle realizes is quite possiblethe agent is prevented from tasting the food.

Affirming the conclusion of a practical syllogism seems an essentially cognitive act (even if it normally has a motivational element); it is roughly an endorsing of a proposition, whereas decision, by virtue of being a deliberative desire and apparently entailing an intention to do the thing decided on, is essentially motivational. Nevertheless, clearly Aristotle is thinking of the relevant kind of judgmentroughly, that one must (or ought, all things considered) do somethingas normally implying a decision to do it.

Aristotle on practical reasoning and the structure of action 13

In part, the connection between the two kinds of decision might be expressed in terms of a relation between cognitive and behavioral decisions. Cognitive decision, decision that, which is the kind one might identify with the concluding judgment, tends to produce a behavioral decision, a decision to do, particularly insofar as the latter decision is understood as a deliberative desire. This is not to say that Aristotle employed the relevant term in these two ways; but it is noteworthy that it has both functions. The cognitive function is appropriate to decisions playing the role of emerging from (and even expressing the conclusion of) reasoning; the behavioral function makes it practical, either in the direct sense that it is itself action or in the indirect sense that it expresses at once a reason for acting and some degree of motivation to act.

If this reading is correct, then the kind of reasoning that normally instantiates a practical syllogism is practical reasoning in the broad sense of reasoning that concludes with an answer to a practical question, such as, paradigmatically, What am I to do? asked in the context of a felt problem. If my problem is to cure the patient and I deliberate toward that end, then upon concluding (cognitively deciding, one might say) that I should reach for the tablets to my left, I have arrived at an answer to my problem. I then solve the problem by reaching for the tablets and giving them to the patient. Since Aristotle took the action constituting a solution to occur at once given the agents unimpeded ability to do the crucial thing, he sometimes talked as if the action itself were the conclusion. But I do not believe we must take the text to assert precisely this.9 It is noteworthy that he speaks of the physicians deliberating about what he will do rather than simply about what to do. This may suggest that the concluding element in the reasoning is a cognitive item (roughly, one that, like belief, has a truth-valued object and may be called true or false), say a resolution (or judgment or perhaps cognitive decision), to the effect that one will (or should) do something. The decision to do it, which may or may not be behaviorally instantiated by the immediate performance of the action, might then be seen as the appropriate action to be performed in response to the drawing of this conclusion.

3 WEAKNESS OF WILL

If Aristotle took concluding in favor of an action normally to imply deciding to perform it, then we must ask how he allowed for weakness of will, in the sense of acting (normally intentionally) against ones better judgment, where this is precisely the sort of judgment which, like This must be tasted, concludes a practical syllogism.10 We might call actions of this kind incontinent for short. In discussing them, I cannot present his overall account of weakness of will, if indeed he offered a fully unified account. Since my concern is with his view of practical reasoning, I simply want to indicate how he saw practical reasoning as allowing for incontinent action.

From this point of view, the following passage is especially important:

Suppose, then, that someone has (a) the universal belief [say, that sweets are to be avoided], and it hinders him from tasting; he has (b) the second belief, that everything sweet is pleasant and this is sweet, and this belief (b) is active; and he also has appetite. Hence the belief (c) tells him to

Practical reasoning and ethical decision 14

avoid this, but appetite leads him on, since it is capable of moving each of the [bodily] parts.

The result, then, is that in a way reason and belief make him act incontinently. The belief (b) is contrary to correct reason (a), but only coincidentally, not in itself.

(NE: 1147a311147b2)

It looks as if one piece of reasoning presupposed here is prohibitional: from a universal premise, say that (a) sweets are to be avoided, to a negative judgment, say that (c) this is to be avoided, which hinders but does not prevent the tasting. Its minor premise is presumably a partly perceptual one to the effect that this, being sweet, is to be avoided.

Moreover, while we need not suppose that there is a second piece of practical reasoning, there are possible cases in which there is also competing, appetitional reasoning: from the premises that (1) everything sweet is pleasant [to taste] and (2) this is [would taste] sweet, to the conclusion that (3) this must be tasted. But positing a second piece of practical reasoning is apparently not Aristotles way of viewing the case. For then there would be an opinion, namely, (3), opposed to right reason, since reason dictates that things of this kind are not to be tasted.

Furthermore, Aristotle specifically tells us that it is appetite, for instance a ravenous desire for sweets, that opposes right reason. Even if we take this opinion to represent appetite, we still have a problem: how can appetite prevail if a properly drawn practical conclusion, such as that this is to be avoided, at least normally implies acting accordingly when one can? Incontinent agents are not, after all, unable to act rightly. The agent here is simply weak-willed, yet appears to have also reasoned in accordance with the first syllogism, from (say) the premises that sweets are not to be tasted and this is such a sweet, to the conclusion that this is not to be tasted. Thus, even assuming the passage should be read as implying competing syllogisms like the pair I have formulated, there remains the question why the syllogism on the side of right reason does not prevail in action. Indeed, the problem is pressing even if an incontinent action is imagined as opposed to a judgment not arising from practical reasoning. Suppose right reason is represented only by a judgment not (at least at the time) based on any practical reasoning? How can right reason not prevail even then?

Aristotles treatment of the problem drawsfruitfully, I believeon a distinction between kinds of knowledge:

And since the last premise (b) is about something perceptible, and controls action, this must be what the incontinent person does not have when he is being affected. Or rather the way he has it is not knowledge of it, but, as we saw, [merely] saying the words as the drunk says the words of Empedoclesthe knowledge that is present when someone is affected by incontinence, and that is dragged about because he is affected, is not the sort that seems to be knowledge to the full extent [in (c)], but only perceptual knowledge [in (b)].

(NE: 1147b917)

Aristotle on practical reasoning and the structure of action 15

Appetite can overcome ones (partly) perceptual knowledge that this is sweet and not to be tasted, but it cannot overcome what constitutes ones knowledge to the full extent that things of this sort are not to be tasted. It is as if appetite detached the perceptual knowledge that this is sweetwhich Irwin suggests (p. 352) may be what Aristotle referred to as the belief which is activefrom the universal known through right reason, and instead attached it to the object of desire. If there is competing reasoning, that object may be expressed in a universal, say that sweet things must be tasted; if there is not, then appetite may affect action more directly. Still, we may ask, how in either case is it possible for the knowledge that should direct action to be relegated to this ineffectual position?

Here it is essential to consider what Aristotle says about such knowledge. One important point is that

Saying the words that come from knowledge is no sign [of fully having it]. Further, those who have just learnt something do not yet know it, though they string the words together; for it must grow into them, and this needs time.

Hence we must suppose that incontinents say the words in the way actors do.

(NE: 1147a1824)

In these and other passages, Aristotle is distinguishing both different kinds of knowledge and different ways of having it: one reading of his suggested distinction between kinds of knowledgeor perhaps one of the two or more distinctions of kind he had in mindis that it holds between a form of recognitionknowing what, for instance what one is eatingand a kind of convictionknowing that, for example that things of a certain kind are to be avoided.

The former, recognitional knowledge, represented in the minor premise, is in a way overshadowed, and it operates under the control of appetite; the latter convictional knowledge, the conviction of right reason, represented in the major premise, is not. We can act against our better judgment precisely because we do not clearly grasp what we are doing, or at least do not grasp it in the right way. We may realize we are eating cake, or even a sweet, but we at best imperfectly know we are eating something not to be tasted. This may occur even if we initially decided to act on the judgment that we are to avoid tasting this. That point might account for our being initially hindered in acting. It may also be true that even the knowledge representing right reason can in some sense be overcome, since it is not unqualifiedly present in the first place, in the sense that the knowledge is not, at the time, fully possessed.

On this reading of Aristotle, one kind of weakness of will is possible because agents can act against their better judgment (incontinently) when that judgment, or some other factor crucial in the genesis of the action, is, though not forgotten, obscured by appetite or other elements in the situation and so, in some way, inadequately known. Call this an obscured knowledge reading. One possible case here is inadequate knowledge of the major premise, as some of Aristotles examples suggest; and we might speculate that if ones knowledge of that, or of other relevant principles of conduct, were fully adequate, ones knowledge of the minor would not be inadequate.

Practical reasoning and ethical decision 16

In the light of Aristotles remarks about knowledge of the major, we can see another plausible reading of his position on incontinence, which yields a different interpretation: even if the major is in one way unqualifiedly known, it is not fully integrated into ones character. The latter possibility, at least, is illustrated by Aristotles comments, quoted above, about beginning students, who may have just learned something, yet do not (fully) know it, since it must grow into them.11 Thus, incontinent agents may know the major, yet not have integrated that knowledge into their motivational systems, say because they lack the required motivation to act on the knowledge, or lack certain habits, or both. The explanation need not be that they are like novices in the subject matter. Aristotle seems to leave open how deep knowledge can be in us intellectually without being integrated into our motivational systems. It may be that even if incontinent agents do fully know both premises, they can still act incontinently; for their knowledge of the major (and perhaps also of the minor) is not integrated into their character, and they do not use it.

This interpretation is particularly plausible so far as Aristotle stresses using knowledge, as when he says that

[W]e speak of knowing in two ways, and ascribe it to someone who has it without using it and to someone who is using it. Hence it will matter whether someone has the knowledge that his action is wrong, without attending to his knowledge, or both has and attends to itwrong action when he does not attend to his knowledge does not seem extraordinary.

(NE: 1146b315)

Unused knowledge is (at least at the time) unintegrated; on the other hand, knowledge not attended to is especially likely to be unused. When incontinence occurs in cases of unused knowledge, however, it is still intelligible. For since reason and belief make the agent act, there is some kind of reasoning and opinion, a kind that matches practical reasoning in its means-end structure, yet represents only emotion (or appetite), rather than right reason, in its major premise. Indeed, this same passage suggests that obscured knowledge, at least of the content of the particular judgment against which one is acting, can itself account for some kinds of incontinence, even if lack of integrationwhich Aristotle implicitly distinguishes from obscured knowledgedoes not occur.

On the insufficient integration reading, weakness of will can apparently occur despite an awareness of an inconsistency (or incongruity) between ones (incontinent) action and ones practical judgment, since the failure is not in the clarity of ones knowledge but in its integration. Doubtless Aristotle emphasized lack of integration as a factor in at least some cases of incontinence; and the view in question helps to explain why incontinence is naturally called weakness of will, since failures of will are clearest where the agent definitely realizes what reason directs. Butparticularly if we do not take him to be making a radical break with Socrates on this issueit is doubtful whether he would countenance the possibility of such clear-eyed weakness of will.12 Possibly he took failure of integration to occur only when the relevant knowledge is either not fully had or not adequately attended to. In any event, each reading may give us part of the truth: both a deficiency in the quality of the agents knowledge and an inadequacy of its motivational integration (or some other kind of integration in the agent) can provide Aristotelian explanations of how weakness of will is possible.

Aristotle on practical reasoning and the structure of action 17

4 PRACTICAL AND THEORETICAL REASONING

On one interpretation of Aristotle, practical and theoretical reasoning contrast sharply in virtue of having different sorts of conclusions: an action in the first case and, in the second, something quite different (presumably a cognition such as a judgment). Anscombe has plausibly argued for this view. She maintains that

we may accept from Aristotle that practical reasoning is essentially concerned with what is capable of turning out variously, without thinking that this subject matter is enough to make reasoning about it practical. There is a difference of form between reasoning leading to action and reasoning for the truth of a conclusion. Aristotle however liked to stress the similarity between the kinds of reasoning, saying [De Motu Animalium VII] that what happens is the same in both. There are indeed three types of casethe theoretical syllogism and also the idle practical syllogism [NE: 1147a278] which is just a classroom example. In both of these the conclusion is said by the mind which infers it. And there is the practical syllogism proper. Here the conclusion is an action whose point is shewn by the premises, which are now, so to speak, on active service.13

One may wonder, of course, how reasoning, which is normally conceived as having as its conclusion something capable of truth or falsity, can have an action as its conclusion. On the other hand, though actions cannot be true or false, they can be supported by premises, particularly premises that show their point. Let us pursue the action-as-conclusion view further.14

The passage that perhaps most strongly supports the action-as-conclusion interpretation is De Motu Animalium 701a425. Consider this part of it first:

What happens seems parallel to the case of thinking and inferring about the immovable objects of science. There the end is the truth seen (for, when one conceives the two premises, one at once conceives and comprehends the conclusion), but here the two premises result in an actionfor example, one conceives that every man [in this situation] ought to walk, one is a man oneself; straightaway one walks.

(701a814)

Aristotle is here drawing an analogy between what results from conceiving the (accepted) premises of a theoretical argument and what results from conceiving those of a practical one; and clearly he puts acting parallel to conceiving and comprehending the conclusion of ones theoretical premises. But he does not say that the action is the conclusion. Indeed, he leaves open the possibility that the action is the indirect result of conceiving the premises, where the direct result is the same sort of thing we have in the theoretical casenamely, drawing, as ones conclusion (in a sense normally implying accepting), the practical proposition implied by the premises, say that I ought to walk.

If this seems to make the parallel inexact, note that on the assumption that Aristotle distinguished conceiving from believing, he might be read as having left room for

Practical reasoning and ethical decision 18

incontinent belief, as where one conceives a conclusion which, in the light of ones premises, one judges one ought to believe, but, because it is highly distasteful, one does not believe. Granted, in 1920 he goes on to say that the conclusion, the I have to make a cloak, is an action (Martha Nussbaums translation). But it is important that here the conclusion, though called an action, is apparently truth-valued and is not described as something doneunless he is thinking of the drawing of it as the action here referred to or perhaps of what we might call setting oneself to make a cloak. Certainly the idea that drawing the conclusion is the action would fit the parallel he is making, for the drawing of a conclusion, if taken to be like making a judgment or forming a belief, is no less behavioral in theoretical reasoning than in practical reasoning.

This reading is supported by a number of points. One is that it makes good sense of the way Aristotle emphasizes the parallel between the practical and theoretical cases; if the contrast is as strong as the action-as-conclusion view has it, one would expect him to stress the differences both more than the similarities and more than he does. Second, I have already quoted him as saying, of the case where two premises are combined as they are when a universal rule is realized in a particular case, that it is necessary in purely theoretical beliefs for the soul to affirm what has been concluded, and in beliefs about production to act at once on what has been concluded (NE: 1147a278, emphasis added). Here he apparently distinguishes between the drawing (or affirming) of the conclusion and the action that results from drawing it, specifically, is taken on what has been concluded.

One might ask why he does not single out drawing the conclusion as a separate act. Since it is self-evident what the conclusion is in the simple instances imagined, in which a universal role is realized in a particular case, he has no need to mention drawing the conclusion as a separate act. But that the conclusion is distinct from the action seems implicit in his maintaining that the action occurs at once for someone who is able and unhindered (NE: 1147a301). When inability or prevention occurs, apparently the conclusion is drawn, but not acted on.15

Indeed, if it is not possible for the agent to conclude the reasoning without acting on it, then it becomes at least more difficult to account for some of Aristotles points about weakness of will. For in that case it certainly appears that emotion overcomes practical reasoning precisely in the sense that one completes itif without appropriate knowledge of its constituents, particularly its major premiseand fails to act on it. One might argue that in cases of incontinence it is only the premises, not the reasoning as a whole, that the agent fails to act on. But Aristotle does not say this and speaks of the reasoning as if it were completed.16 Often, at least, incontinent action represents not uncompleted practical reasoning, but unsuccessful practical reasoning.

There are, to be sure, problems for the suggested interpretation of Aristotles conception of practical reasoning. One is that, as Anscombe quite rightly brings out, from Aristotles point that practical reasoning concerns what can turn out variously, we must not infer that this restriction of subject matter was his criterion for what constitutes practical reasoning. We can surely do theoretical reasoning about any subject (though where the reasoning is scientificas at least the paradigm cases areit must concern matters that admit of the appropriate necessary connections). Another problem is how to take Aristotles point that the last premise is a belief about something perceptible, and controls action (NE: 1147b910, emphasis added). This point makes it appear that

Aristotle on practical reasoning and the structure of action 19

nothing else, such as drawing a conclusion, controls action (though some commentators have taken he teleutaia protasis here to mean conclusion rather than last premisea reading which, though I am not adopting it, certainly supports my interpretation as against the action-as-conclusion view). Aristotles point here may also seem to leave action as the only candidate for the conclusion. Let us consider these problems in turn.

On the interpretation I have suggested, it is not merely subject matter that accounts for the practicality of practical reasoning. If the reasoning occurs in the context of pursuing an end, and if this pursuit includes commitment to a judgment of what one ought to do, then its conclusion may indeed be expected to lead to ones acting at once provided one is able and unhindered. Thus, practical reasoning may be seen as undertaken for a practical purposeachieving some endand normally as issuing in action. Furthermore, its conclusion is not just any judgment, but a practical one, to the effect that one must do something. On this view of Aristotles conception of practical reasoning, then, we have both a significant contrast between practical and theoretical reasoning and an account of his emphasis on the parallels between them.

The other difficultythat, since the acceptance of the minor premise controls the action, the action is the only conclusion for which there is roomcan be resolved by three points. First, acceptance of the minor premise can determine our action even if it does not do so singly or, more important, directly, that is, without the mediation of some other psychological element. Second, even if the minor (or our conceiving and accepting it) should directly determine action, it does not follow that we do not also in some way draw a propositional conclusion, perhaps as a result of conceiving both premises. Drawing this conclusion might have a guiding role with respect to how the action is carried out even if the minor in some sense plays the genetic determining role.

The third point is more positive. Since perception is crucial in fully conceiving of the minor premise, which characteristically concerns a means one perceptually grasps here and now, it is appropriate that the premise play a crucial role in determining the action: the chain from ones ultimate goal, such as healing the patient, back to oneself is completed at the point where one accepts that premise (at which point one also normally judges in favor of the means it identifies); and given the background motivation that underlies ones reasoning, say the desire to heal ones patient, one now acts if one can and is not prevented. Perception can, then, in a way control action without the actions having to be a conclusion of the reasoning. The action may issue from the reasoning at the same time as the conclusion favoring it; the perception both calls for the conclusion and guides the action.

In my view, then, Aristotle may be emphasizing the determination relation he describes in accounting for incontinence, precisely because, when incontinence occurs, a conclusion is drawnin accord with right reasonon which the agent fails to act, and the failure seems explainable by appeal to a similar determination. One main case is appetites simply overriding the practical judgment, as where ones realization of the minor premisethat this is a sweetis not integrated with ones belief of the major; such incontinence may or may not bespeak impetuosity. Another kind of incontinence involves a competing practical syllogism, say with a major premise to the effect that sweets are delectable, which does prevail in action, presumably in part through the force of the same minor premise. In short, the determination relation is stressed as part of the causal account of incontinent action; it does not preclude, and is not in tension with, the

Practical reasoning and ethical decision 20

view that (normally, at least) practical reasoning has both an obviously implied judgmental conclusion as its terminal element and, typically, an action as its issue.

5 THE EXPLANATION OF ACTION

Clearly, practical reasoning as I have suggested Aristotle conceived it provides a way to understand and explain actions. It makes actions that are based on practical reasoning intelligible as conduct in accordance with practical judgment. These actions, in turn, are based on at least one proposition (the major premise) held as a guiding principle and, in a reasonable person, at least prima facie correct.

Aristotles conception of practical reasoning also yields a causal account of actions based on such reasoning. There are at least two important points here. First, the agent is reasoning in the context of a desired end, at least typically in a way that includes a commitment to some principle; this provides motivation for the action issuing from the reasoning. Second, the minor premise expresses a perceptual grasp here and now of what the agent takes to be a means17 to the end (or to acting in accord with the principle); this provides guidance for the action.

The guidance provided by the minor premise is twofold. It is exercised in part by a belief to the effect that the end can be achieved by a certain kind of action, say reaching for a sweet; and it derives partly from a perceptual event with a kind of causal power that enables the belief to initiate action. The perception starts one off in acting in an appropriate way, for instance reaching into the cabinet; the belief helps to sustain and guide the action, say to keep one searching the shelves until one finds the right container. The whole process may be virtually instantaneous, as where, driving at high speed, we see a tumbleweed in our path and decide to drive over it rather than risk swerving.

If Aristotle takes all our intentional actions, as opposed to merely voluntary ones, to arise from practical reasoning, as he sometimes appears to,18 then his view of practical reasoning provides a good account of how intentional action in general is to be understood and explained. Suppose, however, that he should be read as placing practical reasoning only within the context of deliberation and he also allows for the possibility of intentional action that occurs outside this context, as where, after a concert in which one heard Finlandia, one starts humming its melody for pleasure. Whether or not he is to be so read, there surely seem to be intentional actions that do not arise from practical reasoning,19 and if so it is important to consider how Aristotle might account for them.

The first point to note here is that at least the paradigms of intentional actions are all of a kind that can arise from practical reasoning. If so, then it may be open to Aristotle to hold what we might call the correspondence thesis: the view that to every intentional action there corresponds at least one practical argument whose premises (in some way) express motivation and belief jointly sufficient to explain the action.20 For instance, even if I spontaneously eat an apple because I am hungry, and without doing practical reasoning, I may eat it in order to reduce my hunger, which Aristotle might have taken to imply that, with no change in my motivation or beliefs, I could have reasoned from practical premises to a conclusion favoring eating one, say from the premises that eating apples relieves hunger and that eating this apple before me will relieve my hunger.

Aristotle on practical reasoning and the structure of action 21

6 INTRINSICALLY MOTIVATED ACTION

There is, however, some difficulty with this approach to defending the correspondence thesis. Consider actions performed for their own sake, such as playing tennis for pleasure. Let us start by setting aside an argument which, though unsound, can cause confusion: actions performed (entirely) for their own sake are ends in themselves, and if so they are not such that we can deliberate about them; hence, practical reasoning cannot be directed toward them either. Now even if it should be true that for Aristotle practical reasoning is deliberation, or is possible only for actions subject to deliberation, this argument is defective. For Aristotles view that we do not deliberate about ends plainly permits deliberation about what means to take to an intrinsically motivated action, and in that way we can deliberate about the action.

Aristotles view also allows that an end final in one context, say healing a patient, may be a means in another, say a means to making a living. His position is that every deliberation is relative to some end that governs the context and is not therein a subject of deliberation; but it is only the end final in the context that we cannot deliberate about in that context. Moreover, it is only the absolutely final end, happiness, that, in any context, cannot (or at least cannot rationally) be pursued as a means to something further.

So far so good. But consider action both performed for its own sake and basicbasic in the sense that it is not performed by performing any other action. Not only is it not aimed, in any obvious sense, at anything further; as basic, it is not performed by doing anything else; the deed is done at will. Hence, there appears to be no room to reason either about what means will achieve it or about how to make it a successful means to anything else. We have already seen one Aristotelian response to the latter point; there is simply no need for deliberation to apply to every action qua means to something further, since not all actions are properly performed for a further end. This truth may be misleading, however, unless more is said. There is, for all intentional action, a final endnamely, happinessand even actions performed for their own sake may be argued to be performed directly, even if not self-consciously, in order to realize part of that final end conceived as activity (as, for example, NE: 1098a9a); hence, the action may be conceived as a constitutive means to happiness.

How are we to understand happiness as an end? It is not a further end of an action performed for its own sake, such as a deed done simply as the magnanimous gesture; happiness is the intrinsic end of such an action. While the action itself, qua activity partly constitutive of happiness, might be the only relevant end, it is, ultimately, performed for the happiness it yields (or is expected to yield). This is not to say that it must be pursued under a conception that explicitly links it to happiness, for instance being conducive to my happiness. Aristotle is best read as allowing that it be simply the kind of thing that is partly constitutive of happiness and, on the basis of reflection, may be so viewed and accordingly sought. This allows, however, that it be pursued in some way as a constitutive means, say as honorable.

These points about the sense in which intrinsically motivated actions can be constitutive means to happiness do not indicate how anything can be conceived as a means to them when they are basic. This is not a problem for the correspondence thesis. A proponent of that thesis needs a way to exhibit basic action performed for its own sake as in some way reasoned, but not as done by performing some more basic action. If, for

Practical reasoning and ethical decision 22

all intentional action, there is a corresponding practical argument, then all such action must be in some sense aimed; but this does not imply the possibility of being aimed at by some other action, and for basic action there is no distinct action capable of being so aimed. (Granted, one can do somethingsuch as free ones scheduleto cause oneself to do something basic, such as swim across a pool, but this is not trying to swim by doing something else as a means, as one locks a door by turning a key.) It is not entirely clear, however, how intrinsically motivated actions can even be aimed; they are certainly not aimed at any further action or at causal consequences. Let us explore in more detail how such actions might be conceived in an Aristotelian framework.

Consider humming for pleasure, which is a special case of doing something for its own sake and would not normally seem to correspond to practical reasoning. Although one is not humming for a further end, one is humming because of something about doing so, say the distinctively pleasing sounds. If humming may be viewed as a (constitutive) means of experiencing these sounds, then there might after all be a corresponding practical argument. Its premises might be that these pleasing sounds are to be enjoyed in ones humming (roughly, enjoyably realized in humming), and that humming now (perhaps in a certain manner) constitutes the way to enjoy them. This may seem artificial, but it is intelligible. Indeed, acceptance of the minor premise may readily be seen to yield an Aristotelian decision, manifested in (or indeed identical with) the agents acting straightaway.

The idea here is twofold. First, even when one is doing something for its own sake (and basically), there is something about it, such as its pleasing sound, which one wants. Second, one can believe, without triviality, that doing it (in a certain way) constitutes getting that, if only because it is possible, and perhaps all too easy, to do the thing in question without achieving what one wants, say the pleasing sound. Ones pleasure (or other end sought for its own sake) is constituted by doing the thing in a certain way, not just by doing it. Hence, it is intelligible for the minor premise to connect the action with the intrinsic character that action has when performed in a relevant way. The intrinsic end supervenes on the humming, which is a constitutive means to it; but it does not supervene on just any humming. This is why, even in humming for its own sake, one can be doing so as a means to the supervenient pleasure. The suggested account is consistent with Aristotles view of pleasure,21 so if it is successful then the correspondence thesis can be extended to all intentional action, including intrinsically motivated basic action.

7 THE STRUCTURE OF ACTION

The correspondence thesis rests on the idea that intentional action has a certain structure, whether it actually issues from practical reasoning or not. Even if the correspondence thesis as developed in section 5 does not provide an account of how Aristotle could use his conception of practical reasoning to understand intentional action in general, other views he held about action might accomplish this, and they can also clarify his account of practical reasoning.

As Aristotle conceives action, a central element in understanding it is the good. He first considers what we might call subsidiary goods, the kinds that govern deliberation in

Aristotle on practical reasoning and the structure of action 23

a given sphere, yet are not final in the sense that they could never be sought as a means to something further. He says that in each art the good is

that for the sake of which other things are donein medicine this is health, in generalship victoryin every action and decision it is the end.

(NE: 1097a1821)

But the best good must be something that is complete:

Now happiness more than anything else seems unconditionally complete, since we always [choose it, and also] choose it because of itself, never because of something else.

Honor, pleasure, understanding and every virtue we certainly choose because of themselves, since we would choose each of them even if it had no further result, but also choose them for the sake of happiness, supposing that through them we shall be happy.22

(NE: 1097a28b5)

Happiness, then, in the sense of flourishing, stands as our final unifying end: we may seek other things for their own sake, but only when we believe that through themas constitutive and not merely instrumental means, I take itthey will bring happiness.

More specifically, in a given context, say medicine, we may not have happiness in mind in acting, for example in prescribing a remedy. For here our deliberation and practical reasoning are governed by the end of health. But that end in turn is sustained by the final end, presumably in the sense that our seeking the former end is explained by our desire for happiness, together with our belief that achieving the relevant virtue, say the medical virtue that commits us to healing, conduces to happiness. This explanatory sustaining relation does not require that there be in our minds any inferential chain from our final end to our action, nor even from our main subsidiary end, say to practice healing, to the action. That is, we need not actually infer, in each case of action, from the ultimate premise that happiness is to be achieved, to the intermediate conclusion that (say) health is to be promoted, to the practical judgment that (for instance) I should prescribe this medicine. Given the large number of both our subsidiary ends and our instrumental beliefs, this would create a false, and un-Aristotelian, picture of rational action.

Whether or not there is ever such an inferential chain there, there is always a psychologically less intrusive causal chaina purposive chainthat unifies all our actions in relation to our final end, which, at least in a mature, reflective agent, is their ultimate ground. The existence of such a purposive chain underlying an action implies at least this: if challenged as to why we are doing the thing in question, the chain is always accessible to us as an explanatory device (provided we are both rational and are not, for instance, acting in ignorance). For ordinary purposes, citing a single practical argument may serve to explain an action. But to an agent with sufficient self-knowledge, a full explanatory story is always available.

A natural interpretation of purposive chains is to take each link to be constituted by a practical basis relation analogous to the theoretical basis relation by which one belief is

Practical reasoning and ethical decision 24

grounded in, and in that way based on, a second when the latter expresses an evidential reason for which the former is held. Thus, if I prescribe medicine in order to heal, the former action is based on my wanting (or being in some way motivated) to heal. If I want to heal in order to fulfill my medical obligation, the chain has another link, and my healing is based on my wanting to fulfill my obligation. There will be further links until we come to one anchored in the intrinsic desire for happiness, or at least to an intrinsic desire for something that, like virtuous activity, is part of happiness and can be so conceived on reflection.

The practical basis relation has a special feature that is easily overlooked. It is non-transitive: I can, for instance, heal in order to fulfill my medical obligation and do that in order to contribute to my happiness, even if I am not healing in order to contribute to my happiness. For I need have no belief (or minor premise) to the effect that by healing I will contribute to my happiness. No such connection need be made in every case of intentional action; and if I make none, my action is only indirectly grounded in my desire for happiness, a point that has by no means always been appreciated.23

8 THE ULTIMATE GROUND OF ACTION

The importance of the indirect eudaimonistic grounding of actions, as we might call it, should not be underestimated. It is essential if we are to avoid taking Aristotle to be committed to an implausible psychology of motivation. When an action is so grounded, it still has both a causal link to the desire for happiness and, since that desire sustains it (e.g. via sustaining my wanting to fulfill my obligation), a potential explanatory link to that desire. In short, purposive chains ground all intentional actions in the desire for happiness (or for its constituents), directly orfar more often, it would seemindirectly. (There are counterpart relations between the corresponding abstract entities, for example between the content of my desire to heal and the actiontype, prescribing tablets, a type whose concrete realization is warranted by its capacity to realize that content; but my concern here is mainly with the concrete entities related by singular causal connections.)

For Aristotle, explanation is linked not only to causation, but also to reasonableness. This is in part why he wants to ground our characteristic actions in syllogisms and thereby in right reason. From this point of view, it is not our desire for happiness, but the identity of happiness with the good, that is crucial. Happiness is not only our motivationally final end; it is also the good, and thereby rationally pursued for its own sake. Aristotles reasons for identifying it with the good are rooted in his teleological conception of nature as a realm in which the good of things is constituted by their proper functions. I am not concerned to evaluate his resources for defending this conception. My point is simply that by virtue of purposive chains, the good unifies his normative theory of what constitutes reasonable (including moral) action, as well as his theory of motivation.

Moreover, if it is by a use of reason that we come to know that happiness is the good, then reason is normatively practical, in the sense that it suffices to discover truths expressing at least one non-instrumental evaluative standard of conduct. (Beliefs of such truths may constitute knowledge; I am simply not building truth or knowledge into the basic characterization of reason as normatively practical.) If, in additio